Skip to main content
Version: 0.23.1

Runs Panel – Inspecting Scenario History

The Runs panel is the audit trail for everything that executed as a complete scenario, whether started manually or by the Orchestrator. Runs launched from a selected branch of the action tree are not recorded — only full-scenario executions appear.

note

The panel was called Recent Runs in 0.14.2; the sidebar button is now Runs. The log viewer was rebuilt in 0.23 with column filtering and external-log support.

Why it matters

  • Verify that overnight jobs finished.
  • Pinpoint the exact action, log line and screen where a failure happened.
  • Keep evidence of what the robot did, for audits and for handovers.

Opening the panel

Click the clock icon in the sidebar. The list of past runs appears on one side and the detail of the selected run on the other; with nothing selected the detail pane reads No selected run data. Reload Runs refreshes the list — use it when another Editor instance is executing scenarios.

📸 Screenshot needed — /img/editor/recent-runs/panel_view.png

Capture: the Runs panel with a failed run selected, so the run list, the Details tab and the run's metadata are all visible. Why it must be retaken: the whole log viewer was rebuilt in 0.23 — this is the most outdated screenshot in the documentation.


Run statuses

StatusMeaning
SuccessFinished without errors.
Success - retried after exceptionFinished, but at least one action failed and was retried. Worth investigating — the scenario is fragile even though the run is green.
FailureAn action ended with an error and the scenario stopped.
CancelledStopped deliberately with Stop.
Cancelled - Force quitAborted with Force Stop.
CorruptedThe run log could not be read completely.

Each run also records its Execution ModeManual or Scheduled — which tells you whether a person or the Orchestrator started it.


Tabs

TabWhat you seeTypical use
DetailsScenario Name, Project File, File location, Created time, End time, Duration, Status and Execution Mode.Quick health check and runtime statistics.
ProjectRead-only tree of all actions exactly as in the scenario, each marked with the state it ended in.Locate the precise step that failed.
LogsChronological log with Time, Level and Message columns.Read warnings, errors and values printed during the run.
ScreenshotsGallery of the screen captured through the run.See what the robot actually saw at any point.

Combine all three to understand why a step failed: find the failing action in Project, read its message in Logs, then look at the Screenshots taken around it.


Action states in the Project tab

StateMeaning
SuccessThe action completed.
ErrorThe action failed and stopped the scenario.
Error - skippedThe action failed but had Required success off, so the run continued.
SkippedThe action did not run — Run Action was Disabled, or its environment or failure condition did not apply.
StoppedExecution was interrupted at this action.
Not executedThe run never reached this action.
RunningCurrently executing.
UnknownThe log was written by a newer Editor version than the one reading it.
tip

Error - skipped deserves attention. The run is green overall, but something did fail — scanning for it regularly catches scenarios that have been quietly half-working for weeks.


Filtering and external logs

FeatureHow
Filter a columnRight-click a filterable column header. An active filter is marked with a dot on the header.
Log levelsEntries are classified DEBUG, INFO, WARNING, ERROR. Set Settings → Edit Configuration → Aiviro → Logging Level to DEBUG before reproducing a problem.
Open an external logFile → Open external log.zip log (editor) or log directory (core). This is how you inspect a run from a colleague's machine or a production VM.
Sending a run for support

File → Open external log → .zip log (editor) points at the packaged log for a run. Attach that archive to a support ticket — it carries the log, the action results and the screenshots together.


Investigating a failure

  1. Select the failed run at the top of the list.
  2. In Project, expand the tree until you find the first action marked Error — that is where the trouble started, not necessarily where the run stopped.
  3. Switch to Logs and read the message at the matching timestamp.
  4. Open Screenshots and look at the same point. Most failures are one of: the screen was not ready yet (add a Wait For), the element was there but described too loosely (tighten the search object), or an unexpected dialog appeared.
  5. Fix the scenario, re-run, and confirm the entry turns green.